How to Compose New Product Development Portfolios : Optimal Resource Commitment

The process of allocating scarce resources to competing internal and external new product development (NPD) projects comprises two key phases: (i) the collection of relevant information about the projects under consideration and (ii) the allocation of resources to the most promising projects. What makes the management of resource allocation processes challenging is that firms must carefully control the tensions that arise between these two phases. Most notably, a higher level of allocation flexibility in the second phase allows firms to better allocate their resources based on the information they obtain to the most promising projects; however, greater allocation flexibility also induces fiercer competition for resources among the projects, which can severely undermine the reliability of a firm's information acquisition efforts in the first phase and thus deteriorate its overall allocation decisions. Hence, we ask the following question: Which resource allocation policies can best support a firm in coordinating its (ex ante) information acquisition efforts with an optimal degree of resource commitment when the firm contemplates investing in internal and external NPD projects? We study this question by developing a principal-multiagent model, and we identify different contingencies that strongly impact a firm's optimal choice of resource allocation policy, including the projects' market potential and riskiness, the quality of acquired information, and the severity of agency issues. We find that increasing resource allocation flexibility is not always a prudent strategy and that maintaining a combination of dedicated and flexible resource buckets may be an effective way for firms to limit their cost of information acquisition without compromising the scope and composition of their NPD portfolios. We complement these foundational insights with recommendations on how to provide appropriate incentives during the resource allocation process